The present disclosure relates to question-answering technology, and more specifically, to systems and methods for creating patient case summaries from electronic medical records (EMRs) using question-answering technology.
In the medical domain, a patient's medical conditions and treatment history are often stored in collection referred to as an electronic medical record (EMR). An EMR may include large volumes of plain text clinician notes, procedures performed along with results, structured data such as medications, and images such as CT scan images. An EMR may contain both structured and unstructured data resulting from a variety of medical care encounters of a patient with medical care providers. Examples of information that may be contained in an EMR include demographic information, family and social history, reports from care providers, procedures undergone, medications prescribed, diagnostic test results, vital signs, and administrative reports. Therefore, EMRs tend to be large; for example, an EMR of a single patient may contain hundreds of megabytes of structured and unstructured content. An aggregated clinical summary of patient information in the EMR is of high value in patient care if it shows the most important points relevant to patient care.